Bumped die and other bumped devices are widely used throughout the electronics industry. As the drive toward smaller electronics continues, the pitch (or spacing) of solder bumps on such bumped devices continues to decrease. The increasingly finer pitches of the solder bumps on bumped die and bumped devices raise concerns about the reliability of these devices. These concerns are being addressed by testing.
A die (or chip) is typically tested during the manufacturing process to ensure that the die conforms to operational specifications. Solder bumps (or balls) are then formed on bond pads of the die using a solder deposition device, such as a solder ball bumper. The solder bumps are typically formed with a height of from 25 μm to 75 μm. The bumped die are then tested by placing conductive test leads in contact with the solder bumps on the die, applying a test signal to the bumps via the test leads, and determining whether the bumped die responds with the proper output signals. If the bumped die tests successfully, it may be installed on a printed circuit board, a chip scale package, a semiconductor module, or other electronics device.
FIG. 1 is a cross-sectional view of a bumped die 10 engaged with a test carrier 20 in accordance with the prior art. In this typical arrangement, the bumped die 10 includes a substrate 12 with a plurality of bond pads 14 thereon. A solder bump 16 (or other suitable conductive material) is formed on each of the bond pads 14. The test carrier 20 has a plurality of contact pads 22 thereon, each of the contact pads 22 being electrically coupled with a test lead 24. For testing of the bumped die 10, the solder bumps 16 engage the contact pads 22 of the test carrier 20, and the appropriate test signals are applied to the bumped die 10 through some of the test leads 24. Output signals from the bumped die 10 are monitored through other test leads 24 to determine whether the bumped die 10 is functioning to specifications. Test carrier apparatus of the type shown in FIG. 1 for testing unpackaged die are described in U.S. Pat. No. 5,519,332 to Wood et. al., incorporated herein by reference.
Testing of the bumped die 10 generally includes four levels of testing. A first or “standard probe” level includes the standard tests for gross functionality of die circuitry. A second or “speed probe” level includes testing the speed performance of the die for the fastest speed grades. A third or “burn-in die” level involves thermal cycling tests intended to drive contaminants into the active circuitry and to detect early failures. And a fourth or “known good die (KGD)” level includes testing to provide a reliability suitable for final products.
To ensure proper transmission of the test signals and output signals, the solder bumps 16 may be temporarily connected with the contact pads 22 by reflowing the bumps, thereby soldering the bumps to the contact pads. After the testing is complete, the solder bumps 16 may be reflowed to disconnect the bumps from the contact pads. Connecting and disconnecting the solder bumps 16 from the contact pads 22, however, involve time consuming processes and may damage the solder bumps 16 or the contact pads 22.
Another problem with soldering the solder bumps 16 to the contact pads 22 is that the coefficient of thermal expansion (CTE) of the bumped die 10 may be appreciably different from the CTE of the test carrier 20. During burn-in die testing, the bumped die 10 and test carrier 20 are placed in a burn-in oven and subjected to temperature cycling (e.g. −55° C. to 150° C.) for a time period of from several minutes to several hours or more. Due to the different CTE of the bumped die 10 and the test carrier 20 and the rigidity of the solder connections, significant stresses may develop throughout the components. These stresses may result in delamination or other damage to the bumped die 16 or the test carrier 20, and may degrade or damage the connection between the solder bumps 16 and the bond pads 14.
An alternate approach to soldering is to simply compress the solder bumps 16 into engagement with the contact pads 22. Ideally, only a small compression force is needed to engage the solder balls 16 against the contact pads 22 so that tests may be conducted. Methods and apparatus for testing die in this manner are fully described in U.S. Pat. No. 5,634,267 to Farnworth and Wood, incorporated herein by reference. The applied compression force, however, must be kept to a minimum because larger forces may damage the circuitry of the bumped die 10 or the test carrier 20
A problem common to both the solder reflow and the compression force methods of engagement is that the solder bumps 16 are not uniformly shaped. As shown in FIG. 1, the solder bumps 16 are usually of different heights. Using typical manufacturing methods and solders, the nominal variation between the tallest and shortest bumps (shown as a distance d on FIG. 1) is presently about 10% of the average solder ball height. Therefore, when the bumped die 10 is placed on the test carrier 20, the shorter solder bumps may not touch the corresponding contact pads. In some cases, especially for very fine pitch solder bumps, the gaps between the shorter solder bumps and the contact pads may be too large to overcome using solder reflow (because of the small volume of solder in each bump) or by using compression force (because of possible damage to the bumped die).
The variation in solder bump height also creates uncertainty in the final assembly of electronics components that include bumped devices. As the number of bumps on the bumped device increases, the failure rate of the assembled package increases due to solder bump non-uniformity.
FIG. 2 is a partial cross-sectional view of the bumped die 10 of FIG. 1 engaged with another conventional test carrier 40. The test carrier 40 includes a test substrate 42 having a plurality of pockets 44 disposed therein. As shown in FIG. 2, the pockets 44 have sloping sidewalls 46, and a pair of contact blades 48 project from opposing sidewalls 46 into each pocket 44. Conductive test leads 50 are formed on the test substrate 42, including on the sidewalls 46 and contact blades 48 of the pockets 44.
During testing, the solder bumps 16 at least partially engage the pockets 44 of the test carrier 40 with the sharp contact blades 48 partially penetrating the solder bumps 16. The solder bumps 16 may also contact the sloping sidewalls 46 of the test carrier 40. Thus, the desired electrical connection between the solder bumps 16 and the test leads 50 may be achieved despite the variation in the solder bump height.
Although the test carrier 40 having pockets 44 with contact blades 48 addresses solder bump height variation, testing solder bumps with the test carrier 40 has several disadvantages. For example, because the contact blades 48 penetrate the solder bumps 16, the solder bumps may be cracked, chipped, or otherwise damaged by the contact blades. The solder bumps 16 may also become stuck to the contact blades 48, requiring additional time and effort to disengage the bumped die 10 from the test carrier 40. Furthermore, the test carrier 40 with the plurality of pockets 44 is relatively costly to fabricate and more difficult to maintain than alternative test carriers having flat contact pads.
FIG. 3 is a partial cross-sectional view of the bumped die 10 of FIG. 1 engaged with another prior art test carrier 60. In this example, the test carrier 60 includes a test substrate 62 having a plurality of pedestals 64 formed thereon. Test leads 66 are disposed on the test substrate 62, each test lead 66 terminating in a contact pad 68 on the top of each pedestal 64. A plurality of projections 69 project from each contact pad 68. Apparatus for testing semiconductor circuitry of the type shown in FIG. 3 are more fully described in U.S. Pat. No. 5,326,428 to Farnworth et. al., U.S. Pat. No. 5,419,807 to Akram and Farnworth, and U.S. Pat. No. 5,483,741 to Akram et. al., which are incorporated herein by reference.
To conduct a test of the bumped die 10, the solder bumps 16 engage the contact pads 68 so that the sharp projections 69 at least partially penetrate the solder bumps 16. The projections 69 may be properly sized to penetrate into the taller solder bumps, allowing the shorter solder bumps to at least contact the projections of the corresponding contact pad 68.
One of the drawbacks of testing bumped die using the carrier 60 having projections 69 is that the projections (like the contact blades 48 described above) may damage the solder bumps 16. Furthermore, the projections 69 are relatively expensive to manufacture, particularly when the projections must be sized to account for a nominal 10% variation in the solder bump height.